But that's where it stopped working. The closer intervals did not follow the natural harmonic series. At dinner tonight I was talking to my dad, and he said he remembered vaguely something having to do with the 12-th root of two. So, when I got back, I tried:
Code:
440*2^(n/12)
and sure enough, it worked.
Code:
nDegreeFreq. (Hz)
0Root440
11/2 stp466.16
2whl stp493.88
3min3rd523.25
4maj3rd554.37
54th587.33
6Tritone622.25
75th659.26
8min6th698.46
9maj6th739.99
10flat 7783.99
117th830.61
12Octave880
So, why exactly 12 was chosen, I don't know, but maybe that's enough info to get someone going on figuring it out. I suspect it may be because it brought the natural intervals to within "aural rounding error" of the natural series.
And that is how I spend my Sunday afternoon.
(I wonder how long till this gets sent over to AO.)
Nice work Noleli2
As for the use of 12... there are 12 semitones per octave - you sort of figured that out yourself!
I'm *defineitely* not an expert on tuning temperaments or just intonation, but I think that the current Western tuning system also has a bit to do with summation and difference tones interacting, so that notes can be in harmony with each other when certain intervals are played.
I think that is why the major 3rd is pitched slightly flatter(?) than it would be if we used just intonation. Somebody help me out here.
^ That comment about dancers not being concerned with counting as accurately as musicians may be true - but I'm sort of at a loss to really understand how a dancer - for whom the interpretation of rhythm and melody should be a rudiment of their craft - can mistake TAKE 5 for a waltz.
I mean, c'mon - it's called TAKE 5 not "Waltz for half a bar and then waltz more quickly for the other half of the bar!"
[EDIT: typos - hate 'em]
You'd be surprised how many dancers can't count to save themselves. They go entirely by rhythm and feel. Drives me nuts. I've even done Tap classes with master teachers who, when asked for the counts (as opposed to a lot of da da da's), reply they don't do counts and that I shouldn't need them anyway (but I do, I do!).
I did a Jazz class a while back where we were learning a dance to Reefer's (You Make Me Feel) Like a Natural Woman. It was only when the teacher's MD who was doing the class also pointed out that it was actually in 6/4 that it dawned on any of us (she was counting it as a slow 4/4).
Classical Ballet often counts 3/4 as 4/4 and that's before you even get into the whole thing of dancing "across" the music. Although admittedly none of those examples are quite in the same ballpark as counting Take 5 as a waltz. But in terms of the choreography I did to Take 5, it followed a waltz-type pattern in terms of the steps themselves and this is what I think Amorph's dancer probably did without realizing her counts weren't of equal value.
As far as the title goes, people not familiar with Brubeck might not realize the little puns he makes when he names pieces.
But I've seen musos do this sort of fudging too. If you're familiar with Cattle and Cane by The Go-Betweens it's 11/4 and to me the pattern is clearly 5/4 + 6/4 but I'm yet to see it notated on the Net where it's not described as 4/4 + 4/4 + 3/4. But that's not the way The Go-Betweens played it.
Jumping ahead.....I hope they don't move it Noleli2. Lots of interesting stuff here that I'm sure is beneficial to many Garageband devotees (as well as others ). I'm lovin' it anyway and look forward to more illuminating posts from you guys. Great stuff!
I am fairly guilty of not counting, I feel my way through songs and rely heavily on my ear. While this isn't a bad thing per se, since in doing this I have gained a good ear and sense of music in the more... intangible sense. It's not so great in the more technical side of things.
Being mostly Self Taught is the primary cause, but I'm learning now!
I'm *defineitely* not an expert on tuning temperaments or just intonation, but I think that the current Western tuning system also has a bit to do with summation and difference tones interacting, so that notes can be in harmony with each other when certain intervals are played.
I think that is why the major 3rd is pitched slightly flatter(?) than it would be if we used just intonation. Somebody help me out here.
Yeah, it's detuned slightly. I think it's because, like I said a few posts back, if you're playing two notes at once, you're obviously going to end up summing those signals. If you think of each of the tones you're playing as sinusoids, and you have twelve exactly equal semitones, then the two periods you're dealing with are commensurable, so adding them will just result in another periodic signal. If they're shifted slightly, you'll end up with a much more complex harmonic structure, which is likely more harmonic, and those upper harmonics will also be much less harsh than any weird summations that might occur in the upper harmonics with a chord with more than two notes in it, if you have the twelve equal divisions but not pure sinusoids.
Maybe tomorrow after class I'll run some MATLAB simulations. Maybe.
[B ]indeed, Music is so wonderful on so many levels. [/B]
Mmmm. I have this theory (one of many!) that there's a whole lot more to the music thang than we really understand. All that relationship between music and mathematics as per the equal temperament stuff. Throw in a bit of Close Encounters of the Third Kind and pretty soon you got music as the salvation that's going to save us all from hellfire and damnation.
And maybe that's all I should say before I get myself declared certifiably crazy.
Mind you Pythagoras and his lot seemed to kind of see the world in these terms.
OK - I've gone back to some old notes of mine and this is what I found:
The equal tempered scale is divided into 12 equal parts (as discovered above). Mathematically this is the 12th root of 2 (let's say: 1.059463).
To get the next semitone you multiply the previous harmonic by the 12th root of 2 (b/c that is equal temperament).
For example: to find A# (a semitone above A 440Hz) is 440 * 1.059643 = 460.143 Hz (which is the fundamental frequency of the note A# (or Bb) above A 440Hz.
Now we need to talk about the overtone series:
Each note has a series of overtones which gives it its timbre (or character).
Basically it builds on the fundamental: octave, fifth, fourth, maj 3rd, min 3rd, min 3rd, tone, tone, tone, tone, semitone etc.
So the note C is really comprised of these other frequencies (at lesser degrees)
C - C' - G' - C" - E" - G" - Bb" - C"' - D"' - E"' - F#"' - G"' etc.
The ratios between these frequencies in the overtone series help determine our intervals.
a Maj 3rd (C to E, for example) is the ratio: 4:5 and 8:10
a min 3rd (E to G, for example) is the ratio 5:6 and 10:12
Now, using the harmonic series we can work out the frequencies of other notes. So, let's say we have a LOW C (65.406Hz) and we want to work out the frequency of A# above A 440 Hz - the same note as our first calculation above using equal temperament. (Remember that A# is the same as Bb)
Well - A# (Bb) is the 7th in the series above, so the fundamental * 7 gives us the frequency of the note we're looking for.
Let's check it out: C (65.406) * 7 = 457.842 Hz. This is the frequency of A# above A 440Hz.
The eagle eyed amongst you would have noticed that in equal temperament the frequency was 460.143Hz, but using the overtone series of ratios the same note had a frequency of 457.842 Hz.
*I am not an expert on this stuff - just putting it out there for those that are interested*
Code:
INTERVALovertone ratios
Unison 1:1 (1/1)
Octave 1:2 (2/1, 6/3, 8/4)
Fifth 2:3 (3/2, 6/4)
Fourth 3:4 (4/3, 8,6)
Maj 3rd 4:5 (5/4, 10/8 )
min 3rd 5:6 (6/5, 12/10)
Maj 6th 3:5 (5/3, 10/6)
min 6th 5:8 (8/5, 16/10)
tone 8:9 or 9:10 - a tone can have two different ratios
semitone 15:16 (16/15)
Now, for a comparison between tuning systems
Code:
Scale note Pythagorean | Just intonation
1 - C -> 1:1 | 1:1
2 - D -> 8:9 | 8:9
3 - E -> 64:81 | 4:5
4 - F -> 3:4 | 3:4
5 - G -> 2:3 | 2:3
6 - A -> 16:27 | 3:5
7 - B -> 243:128 | 8:15
8 - C -> 1:2 | 1:2
Note how the 3rd and 6th in the Pythagorean tuning ratios are flatter than the just intonation system. Again, this has to do with the blending of tones and creating an equal temperament for all keys that is pleasing to our Westernised ears. As I mentioned (and Neloli2 also posted): this has to do with summation and difference tones/frequencies (or Heterodyne components - which is an extension of this topic... for another day perhaps!)
Did any of that make sense?
PS - somebody please help me make these tables look neat - I can't do it and they are SOOO UGLY!!
^ That comment about dancers not being concerned with counting as accurately as musicians may be true - but I'm sort of at a loss to really understand how a dancer - for whom the interpretation of rhythm and melody should be a rudiment of their craft - can mistake TAKE 5 for a waltz.
I can understand why. It has a waltzy lilt to it.
Quote:
I mean, c'mon - it's called TAKE 5 not "Waltz for half a bar and then waltz more quickly for the other half of the bar!"
Yeah, but "Take Five" could be just a phrase. After all, "Three To Get Ready" isn't in three (well, it isn't just in three...), and "Kathy's Waltz" isn't (just) a waltz.
The absolutely brilliant thing about "Time Out" is that if you're not paying attention it just sounds like breezy, catchy cool jazz. People who don't know the first thing about music can hum along. It's when you start trying to figure out just what the hell they're doing that all this incredible complexity reveals itself.
Yeah - I'll give you that... I think I was just wearing my pedantic muso hat when I made that comment.
Quote:
Originally posted by Amorph
Yeah, but "Take Five" could be just a phrase. After all, "Three To Get Ready" isn't in three (well, it isn't just in three...), and "Kathy's Waltz" isn't (just) a waltz.
that was just me being a smart a*se
Quote:
Originally posted by Amorph
The absolutely brilliant thing about "Time Out" is that if you're not paying attention it just sounds like breezy, catchy cool jazz. People who don't know the first thing about music can hum along. It's when you start trying to figure out just what the hell they're doing that all this incredible complexity reveals itself.
Absolutely - that's the great thing about all good art I guess... on the surface it is accessible to the audience on their own individual level - but when they go deeper into it, the rewards are there also.
I'm sure the same could be argued for the coding and UI design of a certain computer manufacturer's OS as well.
But I've seen musos do this sort of fudging too. If you're familiar with Cattle and Cane by The Go-Betweens it's 11/4 and to me the pattern is clearly 5/4 + 6/4 but I'm yet to see it notated on the Net where it's not described as 4/4 + 4/4 + 3/4. But that's not the way The Go-Betweens played it. ...[snip]...
Good pick up craze.
Also, if anybody else is interested in odd metered music check out some Zappa or more recently, Dream Theater. and LTE. Some of this stuff still baffles me! I though i could count - but I'm sure I'm missing something on a couple of their tunes!
You'd be surprised how many dancers can't count to save themselves.
Muahahahahaha...
Warning, joke ahead:
A drummer was asked to accompany a dancer for a piece. So the dancer tells him to play in 3. The drummer starts playing, and the dancer starts dancing, and then she looks at him oddly and stops.
"No, play it in three."
"I was playing in three."
"Well, let's try again."
So the drummer starts playing, and the dancer starts dancing, and the whole thing falls apart again. The dancer turns on the drummer. "I told you to play in three!"
"I am playing in three!"
"No... here, let me count it for you: 1, 2, 3, <pause>, 1, 2, 3..."
I am interested too. I know a little bit about this stuff, how classical temperment was C=256hz and A=432hz. and I know that there was some weird periods when notes would shift around hz. values for whatever reason. But I don't really know that much other than that.
Why we have a 12 tone equal tempered scale has always eluded me, every description I've read has been way over my head
The reason is because if you use other systems, then standard chords (e.g. an F# major triad) will sound very different (out of tune) in relation to other chords, such as a C major triad. This translates to keys whose primary triads (I, IV and V) use such out-of-tune chords - keys such as F# major were avoided before equal temperament for this reason. It's the goal of the equal tempered system that a piece of music can be written for a keyboard instrument in any key and it will sound OK.
Vocalists and other instrumentalists such as string players can easily adjust their tuning on-the-fly, and in fact that's what they do intuitively all the time. However keyboard and fretted instruments are another story, which is why the equal tempered tuning system, where the ratio between each consecutive semitone (half-step) is the same, is our happy compromise.
Still, it's interesting that keys do have certain 'feels' to them. Shouldn't do, since under the equal tempered system they should all sound equal (the ratio of Hz between the notes should be similar). However it should be no surprise that a piece like the Barber Adagio for Strings is written in a mode based around Ab minor (from memory), which is a very warm sounding key. Keys such as D Major have a very bright feel to them in general.
Oh, and for that other person, "Mars" from The Planets is actually in 5/4, not 5/8 FWIW.
Comments
Originally posted by Noleli2
...[snip]...
But that's where it stopped working. The closer intervals did not follow the natural harmonic series. At dinner tonight I was talking to my dad, and he said he remembered vaguely something having to do with the 12-th root of two. So, when I got back, I tried:
Code:
440*2^(n/12)
and sure enough, it worked.
nDegreeFreq. (Hz)
0Root440
11/2 stp466.16
2whl stp493.88
3min3rd523.25
4maj3rd554.37
54th587.33
6Tritone622.25
75th659.26
8min6th698.46
9maj6th739.99
10flat 7783.99
117th830.61
12Octave880
So, why exactly 12 was chosen, I don't know, but maybe that's enough info to get someone going on figuring it out. I suspect it may be because it brought the natural intervals to within "aural rounding error" of the natural series.
And that is how I spend my Sunday afternoon.
(I wonder how long till this gets sent over to AO.)
Nice work Noleli2
As for the use of 12... there are 12 semitones per octave - you sort of figured that out yourself!
I think that is why the major 3rd is pitched slightly flatter(?) than it would be if we used just intonation. Somebody help me out here.
Originally posted by Mac+
^ That comment about dancers not being concerned with counting as accurately as musicians may be true - but I'm sort of at a loss to really understand how a dancer - for whom the interpretation of rhythm and melody should be a rudiment of their craft - can mistake TAKE 5 for a waltz.
I mean, c'mon - it's called TAKE 5 not "Waltz for half a bar and then waltz more quickly for the other half of the bar!"
[EDIT: typos - hate 'em]
You'd be surprised how many dancers can't count to save themselves. They go entirely by rhythm and feel. Drives me nuts. I've even done Tap classes with master teachers who, when asked for the counts (as opposed to a lot of da da da's), reply they don't do counts and that I shouldn't need them anyway (but I do, I do!).
I did a Jazz class a while back where we were learning a dance to Reefer's (You Make Me Feel) Like a Natural Woman. It was only when the teacher's MD who was doing the class also pointed out that it was actually in 6/4 that it dawned on any of us (she was counting it as a slow 4/4).
Classical Ballet often counts 3/4 as 4/4 and that's before you even get into the whole thing of dancing "across" the music. Although admittedly none of those examples are quite in the same ballpark as counting Take 5 as a waltz. But in terms of the choreography I did to Take 5, it followed a waltz-type pattern in terms of the steps themselves and this is what I think Amorph's dancer probably did without realizing her counts weren't of equal value.
As far as the title goes, people not familiar with Brubeck might not realize the little puns he makes when he names pieces.
But I've seen musos do this sort of fudging too. If you're familiar with Cattle and Cane by The Go-Betweens it's 11/4 and to me the pattern is clearly 5/4 + 6/4 but I'm yet to see it notated on the Net where it's not described as 4/4 + 4/4 + 3/4. But that's not the way The Go-Betweens played it.
Jumping ahead.....I hope they don't move it Noleli2. Lots of interesting stuff here that I'm sure is beneficial to many Garageband devotees (as well as others
I am fairly guilty of not counting, I feel my way through songs and rely heavily on my ear. While this isn't a bad thing per se, since in doing this I have gained a good ear and sense of music in the more... intangible sense. It's not so great in the more technical side of things.
Being mostly Self Taught is the primary cause, but I'm learning now!
I'm *defineitely* not an expert on tuning temperaments or just intonation, but I think that the current Western tuning system also has a bit to do with summation and difference tones interacting, so that notes can be in harmony with each other when certain intervals are played.
I think that is why the major 3rd is pitched slightly flatter(?) than it would be if we used just intonation. Somebody help me out here.
Yeah, it's detuned slightly. I think it's because, like I said a few posts back, if you're playing two notes at once, you're obviously going to end up summing those signals. If you think of each of the tones you're playing as sinusoids, and you have twelve exactly equal semitones, then the two periods you're dealing with are commensurable, so adding them will just result in another periodic signal. If they're shifted slightly, you'll end up with a much more complex harmonic structure, which is likely more harmonic, and those upper harmonics will also be much less harsh than any weird summations that might occur in the upper harmonics with a chord with more than two notes in it, if you have the twelve equal divisions but not pure sinusoids.
Maybe tomorrow after class I'll run some MATLAB simulations. Maybe.
Originally posted by Wrong Robot
[B ]indeed, Music is so wonderful on so many levels. [/B]
Mmmm. I have this theory (one of many!) that there's a whole lot more to the music thang than we really understand. All that relationship between music and mathematics as per the equal temperament stuff. Throw in a bit of Close Encounters of the Third Kind and pretty soon you got music as the salvation that's going to save us all from hellfire and damnation.
And maybe that's all I should say before I get myself declared certifiably crazy.
Mind you Pythagoras and his lot seemed to kind of see the world in these terms.
Walk like an Egyptian.
Think like an ancient Greek!
The equal tempered scale is divided into 12 equal parts (as discovered above). Mathematically this is the 12th root of 2 (let's say: 1.059463).
To get the next semitone you multiply the previous harmonic by the 12th root of 2 (b/c that is equal temperament).
For example: to find A# (a semitone above A 440Hz) is 440 * 1.059643 = 460.143 Hz (which is the fundamental frequency of the note A# (or Bb) above A 440Hz.
Now we need to talk about the overtone series:
Each note has a series of overtones which gives it its timbre (or character).
Basically it builds on the fundamental: octave, fifth, fourth, maj 3rd, min 3rd, min 3rd, tone, tone, tone, tone, semitone etc.
So the note C is really comprised of these other frequencies (at lesser degrees)
C - C' - G' - C" - E" - G" - Bb" - C"' - D"' - E"' - F#"' - G"' etc.
1 - 2 - 3 - 4 - 5 - 6 - 7 - 8 - 9 -10 - 11 - 12 etc.
The ratios between these frequencies in the overtone series help determine our intervals.
a Maj 3rd (C to E, for example) is the ratio: 4:5 and 8:10
a min 3rd (E to G, for example) is the ratio 5:6 and 10:12
Now, using the harmonic series we can work out the frequencies of other notes. So, let's say we have a LOW C (65.406Hz) and we want to work out the frequency of A# above A 440 Hz - the same note as our first calculation above using equal temperament. (Remember that A# is the same as Bb)
Well - A# (Bb) is the 7th in the series above, so the fundamental * 7 gives us the frequency of the note we're looking for.
Let's check it out: C (65.406) * 7 = 457.842 Hz. This is the frequency of A# above A 440Hz.
The eagle eyed amongst you would have noticed that in equal temperament the frequency was 460.143Hz, but using the overtone series of ratios the same note had a frequency of 457.842 Hz.
NEXT POST: Raio comparisons
[EDIT - typos hate 'em]
*I am not an expert on this stuff - just putting it out there for those that are interested*
INTERVAL overtone ratios
Unison 1:1 (1/1)
Octave 1:2 (2/1, 6/3, 8/4)
Fifth 2:3 (3/2, 6/4)
Fourth 3:4 (4/3, 8,6)
Maj 3rd 4:5 (5/4, 10/8 )
min 3rd 5:6 (6/5, 12/10)
Maj 6th 3:5 (5/3, 10/6)
min 6th 5:8 (8/5, 16/10)
tone 8:9 or 9:10 - a tone can have two different ratios
semitone 15:16 (16/15)
Now, for a comparison between tuning systems
Scale note Pythagorean | Just intonation
1 - C -> 1:1 | 1:1
2 - D -> 8:9 | 8:9
3 - E -> 64:81 | 4:5
4 - F -> 3:4 | 3:4
5 - G -> 2:3 | 2:3
6 - A -> 16:27 | 3:5
7 - B -> 243:128 | 8:15
8 - C -> 1:2 | 1:2
Note how the 3rd and 6th in the Pythagorean tuning ratios are flatter than the just intonation system. Again, this has to do with the blending of tones and creating an equal temperament for all keys that is pleasing to our Westernised ears. As I mentioned (and Neloli2 also posted): this has to do with summation and difference tones/frequencies (or Heterodyne components - which is an extension of this topic... for another day perhaps!)
Did any of that make sense?
PS - somebody please help me make these tables look neat - I can't do it and they are SOOO UGLY!!
[EDIT - Thanks Wrong Robot... I dig!]
if you want to make those charts look nice, format them with the [ code ] function.
Now, for a comparison between tuning systems
Scale note Pythagorean | Just intonation
1 - C -> 1:1 | 1:1
2 - D -> 8:9 | 8:9
3 - E -> 64:81 | 4:5
4 - F -> 3:4 | 3:4
5 - G -> 2:3 | 2:3
6 - A -> 16:27 | 3:5
7 - B -> 243:128 | 8:15
8 - C -> 1:2 | 1:2
dig?
Originally posted by Mac+
^ That comment about dancers not being concerned with counting as accurately as musicians may be true - but I'm sort of at a loss to really understand how a dancer - for whom the interpretation of rhythm and melody should be a rudiment of their craft - can mistake TAKE 5 for a waltz.
I can understand why. It has a waltzy lilt to it.
I mean, c'mon - it's called TAKE 5 not "Waltz for half a bar and then waltz more quickly for the other half of the bar!"
Yeah, but "Take Five" could be just a phrase. After all, "Three To Get Ready" isn't in three (well, it isn't just in three...), and "Kathy's Waltz" isn't (just) a waltz.
The absolutely brilliant thing about "Time Out" is that if you're not paying attention it just sounds like breezy, catchy cool jazz. People who don't know the first thing about music can hum along. It's when you start trying to figure out just what the hell they're doing that all this incredible complexity reveals itself.
Originally posted by Amorph
I can understand why. It has a waltzy lilt to it.
Yeah - I'll give you that... I think I was just wearing my pedantic muso hat when I made that comment.
Originally posted by Amorph
Yeah, but "Take Five" could be just a phrase. After all, "Three To Get Ready" isn't in three (well, it isn't just in three...), and "Kathy's Waltz" isn't (just) a waltz.
that was just me being a smart a*se
Originally posted by Amorph
The absolutely brilliant thing about "Time Out" is that if you're not paying attention it just sounds like breezy, catchy cool jazz. People who don't know the first thing about music can hum along. It's when you start trying to figure out just what the hell they're doing that all this incredible complexity reveals itself.
Absolutely - that's the great thing about all good art I guess... on the surface it is accessible to the audience on their own individual level - but when they go deeper into it, the rewards are there also.
I'm sure the same could be argued for the coding and UI design of a certain computer manufacturer's OS as well.
Originally posted by crazychester
...[snip]...
But I've seen musos do this sort of fudging too. If you're familiar with Cattle and Cane by The Go-Betweens it's 11/4 and to me the pattern is clearly 5/4 + 6/4 but I'm yet to see it notated on the Net where it's not described as 4/4 + 4/4 + 3/4. But that's not the way The Go-Betweens played it. ...[snip]...
Good pick up craze.
Also, if anybody else is interested in odd metered music check out some Zappa or more recently, Dream Theater. and LTE. Some of this stuff still baffles me! I though i could count - but I'm sure I'm missing something on a couple of their tunes!
Q. How many post modern composers (avant gardé musicians) does it take to screw in a lightbulb?
A. 7 ... in the time of 13.
Originally posted by crazychester
You'd be surprised how many dancers can't count to save themselves.
Muahahahahaha...
Warning, joke ahead:
A drummer was asked to accompany a dancer for a piece. So the dancer tells him to play in 3. The drummer starts playing, and the dancer starts dancing, and then she looks at him oddly and stops.
"No, play it in three."
"I was playing in three."
"Well, let's try again."
So the drummer starts playing, and the dancer starts dancing, and the whole thing falls apart again. The dancer turns on the drummer. "I told you to play in three!"
"I am playing in three!"
"No... here, let me count it for you: 1, 2, 3, <pause>, 1, 2, 3..."
Originally posted by Amorph
"No... here, let me count it for you: 1, 2, 3, <pause>, 1, 2, 3..."
Or the story about the person who counted in seven:
one, two, three, four, five, six, se-ven
Originally posted by Wrong Robot
I am interested too. I know a little bit about this stuff, how classical temperment was C=256hz and A=432hz. and I know that there was some weird periods when notes would shift around hz. values for whatever reason. But I don't really know that much other than that.
Why we have a 12 tone equal tempered scale has always eluded me, every description I've read has been way over my head
The reason is because if you use other systems, then standard chords (e.g. an F# major triad) will sound very different (out of tune) in relation to other chords, such as a C major triad. This translates to keys whose primary triads (I, IV and V) use such out-of-tune chords - keys such as F# major were avoided before equal temperament for this reason. It's the goal of the equal tempered system that a piece of music can be written for a keyboard instrument in any key and it will sound OK.
Vocalists and other instrumentalists such as string players can easily adjust their tuning on-the-fly, and in fact that's what they do intuitively all the time. However keyboard and fretted instruments are another story, which is why the equal tempered tuning system, where the ratio between each consecutive semitone (half-step) is the same, is our happy compromise.
Still, it's interesting that keys do have certain 'feels' to them. Shouldn't do, since under the equal tempered system they should all sound equal (the ratio of Hz between the notes should be similar). However it should be no surprise that a piece like the Barber Adagio for Strings is written in a mode based around Ab minor (from memory), which is a very warm sounding key. Keys such as D Major have a very bright feel to them in general.
Oh, and for that other person, "Mars" from The Planets is actually in 5/4, not 5/8 FWIW.